Measurement Considerations for Quantitative Social Science Research Using Social Media Data
Year: 2020
Type: preprint
Abstract: Harnessing social media data for social science research entails creating measures out of the largely unstructured, noisy data that users generate on different platforms. This harnessing, particularly of data at scale, requires using methods developed in computer science. But it also typically requires integrating these methods with assessments of measurement quality along social science criteria -- reliability, validity and unbiasedness. In this paper, we outline measurement issues that arise when using social media data. We show examples of how to construct measures and discuss different measurement considerations and best practices. We conclude with a discussion of ways to accelerate research in this space, highlighting contributions that can be made by both social scientists and computer scientists.
Authors: Jonathan Ladd, Rebecca Ryan, Lisa Singh, Leticia Bode, Ceren Budak, Frederick Conrad, Elizabeth Cooksey, Pamela Davis-Kean, Keenan Dworak-Fisher, Deen Freelon, Daniel Hopkins, J. Bradford Jensen, Ken Kelley, Renee Miller, Zeina Mneimneh, Josh Pasek, Trivellore Raghunathan, Carole Roan Gresenz, Sudeepa Roy, Stuart Soroka, Michael Traugott
Language: English
Subfield: General Social Sciences
Field: Social Sciences
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W4251377189